New E2E Migration (PW) test project

[HaoK]

• Get tests running on playwright/helix

[HaoK]

This seems like a good initial checkpoint

Currently the single test in the project just verifies that it can go to the page with playwright, without verifying any functionality.

So this checkpoint has:

• New Microsoft.AspNetCore.Components.Migrations.E2ETest project
• Uses playwright on linux/osx/windows helix queues

Next iteration will focus on:

• refactoring files to share between Blazor templates
• adding some more real tests that actually do something

[HaoK]

FYI @captainsafia the razor error does seem to happen on CI, this is it right? I won't trigger a new build until you get a chance to confirm that this is indeed the same issue

src\Components\test\testassets\BasicTestApp\BindCasesComponent.razor(22,1131): error RZ9991: (NETCORE_ENGINEERING_TELEMETRY=Build) The attribute names could not be inferred from bind attribute 'bind-value'. Bind attributes should be of the form 'bind' or 'bind-value' along with their corresponding optional parameters like 'bind-value:event', 'bind:format' etc.

[captainsafia]

Yes! This is the same issue as described in #31023.

This is the first time I'm seeing it happen in CI which is super spooky. Is it possible that a change to configuration made here made this issue appear where it previously didn't? Did it show up in a particular commit?

[HaoK]

Honestly I'm pretty sure I've seen this before on ci in my runs, I usually just restart it in the past

[HaoK]

I have a built in tolerance/expectation that the build fails frequently for things unrelated to my changes, so I'm trained to just kick the build a few times until it starts up... kind of like a lawn mower, needs a few good yanks to start...which is prob not ideal, but it is what it is :)

[HaoK]

One bit of info for you, it did happen after I configured helix-matrix to run, so perhaps that caused it to have bad cache data... since now two jobs are doing a built, idk what gets reused on the build machines between jobs or what not between yml...if anything, perhaps its just bad luck... but that's the only change I made, I wanted to run the full helix matrix as part of this PR to make sure the new test project doesn't explode on the other helix queues

[captainsafia]

@HaoK Yeah I see the issue no longer repros after 963373c. It might very well have something to do with cache state on the machine.

[HaoK]

@javiercn I think its worth getting this in with no real tests as a starting point, so reliability shouldn't be an issue

[captainsafia]

No issues with the approach of getting this in to validate.

Left some questions inline.

[captainsafia]

Why is this commented out?

[HaoK]

I wanted to leave the existing test code there since it might aid in the process of migrating this test to playwright apis later. All the commended out code is the selenium test code.